Combat - Thoughts and discussion

Avatar

By Sanglorian 1 Jul 2018 05:09

Champion · 491 comments

Hi folks,

I've been doing a lot of thinking about Tuxemon combat over the last couple of months, and I thought I'd set up this thread to share our thoughts and trigger discussion.

What's changed since last time I suggested how the combat system could look is that since then I tried programming a combat system for myself (Tuxemon Adventures), which has given me a great appreciation for how difficult programming is! But it also means I can be much more specific in outlining a combat system, which I hope will make it much easier for people to implement in Tuxemon.

The first thing I'd like to talk about is ...

Stats

There's been general agreement since the beginning, I think, over six Stats:

HP: How much damage you can take
Melee: How much damage your Melee attacks do
Ranged: How much damage your Ranged attacks do
Armour: How well you resist damage from Melee attacks
Dodge: How well you resist damage from Ranged attacks
Speed: Whether you act before your opponent in battle

There have been a few suggestions for other stats, like Luck/Precision/Veracity (increasing your critical hit rate) and Aim (increasing the Accuracy of your attacks). 

What could go into a Stat

What an individual tuxemon's Stat values actually are could depend on a number of inputs:

* Level
* Species
* Innate talent
* Experiences
* Quality

Whether your level adds to or multiplies your Stat values would affect how unbalanced a fight between tuxemon of different levels is. For example, does a level 50 tuxemon have Stats that are double that of a level 25 tuxemon? Or more than double? Or less than double?

The input from species could be that each tuxemon species has its own Stat values. I actually prefer abstracting this one step further, and saying that each species belongs to one of about 14 Body Shapes, and Body Shape helps determine Stat values.

"Innate talent" would be a random value for each individual, determined upon creation. Two options have been proposed for this:

* Nature Points: Individual tuxemon increase their Stats at different rates as they level up
* Tastes: One Stat is increased by 10% and one is decreased by 10%

"Experiences" would depend on something that the individual tuxemon had done during its lifetime:

* Battle Points: The more you use ranged attacks, the better your Ranged Stat gets, for example.
* Experience: When you beat a tuxemon with a better Stat than you, your own Stat increases. For example, the protagonist's Memnomnom with Speed 20 and Melee 30 faces a wild Spighter with Speed 30 and Melee 25. At the end of the battle, the Memnomnom gets one Speed Experience (its Speed was lower than the Spighter's) but no Melee Experience (its Melee is higher than the Spighter's).

"Quality" would be what power level the tuxemon is. I proposed "Life Stages", where the original stage of a tuxemon family would get all its Stats multiplied by 0.9, and the final stage would get all its Stats multiplied by 1.1. This reflects that we expect a Rockat to be more powerful than a Rockitten, and a Jemuar to be more powerful than either.

Formula

I've been doing some calculations, and I think the formula below works well:

Body Shape Multiplier * (7 + Tuxemon Level)

With Body Shape Multipliers ranging from 4 to 8, that means that tuxemon of the same level could have Stats that are at most twice as high as another. I think this is a good range; you might notice that I previously suggested Body Shape Multipliers between 2 and 6 - I did some maths and I think that three times higher is too much!

I've put my suggestions for Body Shapes and Body Shape Multipliers on the wiki. I'd be really interested to hear what people think - of the idea in general, of the Multipliers that I've suggested, and of the formula.

Expansion

It would be possible to add some mechanism for innate talent, life stages and experiences to the formula I've suggested, but I feel like keeping it as simple as possible is best for now!

Archive
* A framework for the combat system (my most recent proposal)
* Proposed Tuxemon effects
* 111 Tuxemon techniques
* Let's talk about Tuxemon techniques
* Technique suggestions
* How should stats work?

Last edited by Sanglorian (1 Jul 2018 05:14)

Avatar

By Sanglorian 3 Jul 2018 11:01

Champion · 491 comments

Techniques

Techniques are discrete attacks, blocks, tricks and other manoeuvres that tuxemon perform in combat.

How they work

Acquisition
There are two main options for how techniques are acquired:

  • A tuxemon species knows a certain number of techniques automatically, and they might be able to learn or acquire others from another list.
  • A tuxemon species learns certain techniques at certain levels, and they might be able to learn or acquire others from another list.

Availability

To keep things managable, tuxemon should probably only have access to a finite set of techniques in combat. There are two options:

  • When a tuxemon learns a new technique but they are already at their limit, they have to forget an old technique
  • Out of battle, a tuxemon trainer can choose their tuxemon's techniques, from the list of all techniques that they know

The latter is how EvoCreo does it. However, there is some concern that this takes an interesting decision away from the player.

Additional
Tuxemon could learn additional techniques from a number of sources:

  • Technical Manuals
  • Other tuxemon

Both options are attractive, and having both lets comparisons be made between them: pay a corporation for a Technical Manual, or share knowledge between tuxemon for no charge.

Uses
There are also a number of ways of pacing how many times a technique can be used:

  • Each technique could have a fixed number of Power Points, determining how many times it can be used before a rest is required.
  • Techniques could each have a cost in mana - once mana is too low, they are not available for use. Tuxemon might recover mana over time
  • Techniques could have a recharge time: once used, it takes them a while to be available for use again.
  • Techniques could be used any number of times, perhaps because there is a scissors-paper-rock dynamic where over-using a particular technique is dangerous.
  • Techniques could be rationed in another way, like Dynamons World's technique card deck.

I think recharging is the best way of doing it.

Varieties
There are a number of types of technique you can have:

  • Distinguishing between attacks based on how they use Stats - Melee (the attacking tuxemon bodily strikes the target in close quarters) and Ranged (the attacking tuxemon using a projectile, energy beam, pseudopod or other remote attack)
  • Having a mega technique and a healing technique in dedicated slots, like EvoCreo does
  • Additional generic techniques, like Defend

Elements
The most obvious is for techniques to have Types that match the types that monsters can have, including having two types if such a thing is possible for monsters.

However, there are a few other options:

  • Only being able to have one type.
  • An additional type, "Aether", that takes on the user tuxemon's type(s). This allows for generic attacks like tackles and bites to not belong to a particular type.
  • Special techniques - which don't involve attacks - don't need a type at all.
  • Techniques could have completely different types to the types that monsters have. See Types for more discussion.

Options
There could be any number of options for techniques. Some are below under Other options.

Avatar

By ShadowApex 3 Jul 2018 20:42

Lead Developer · 374 comments
Sanglorian wrote

Uses
There are also a number of ways of pacing how many times a technique can be used:

Another option for uses could be that they have a number of of uses per encounter. For example, during a single battle a certain technique could have 5 uses, and then is refreshed when you enter a new battle.


Avatar

By Sanglorian 6 Jul 2018 12:26

Champion · 491 comments

Capturing

Purpose
Having capturing as part of combat adds a few features to the game:

1. Even battles against weak tuxemon are challenging, since you need to lower their hit points without eliminating them
2. Other tactics become preferable, like using conditions or using techniques that aren't so damaging
3. Capturing comes with a monetary cost, and being sloppy with capturing carries a higher cost
4. Customisation - in a game with different types of Capture Device, you might want to catch your tuxemon in one particular one, or have a collection

How It Works

The main method of capturing that has been proposed is as follows:

1. You buy or otherwise acquire a number of Capture Devices
2. In battle, you weaken the tuxemon by damaging it, and potentially also by giving it a negative condition
3. Instead of taking another action in a Round like using a technique, you use a Capture Device
4. It has a chance of succeeding. If it does, you catch the tuxemon. If it doesn't, then your Turn for the Round is over. You can try again later, however.

There are a number of factors that could be involved in catch likelihood:

* The item used
* The level of the tuxemon
* The HP of the tuxemon (current and maximum)
* The species of tuxemon (some may be harder or easier to catch)
* Conditions affecting the tuxemon
* Some property of the trainer, like if they are well-liked by their other tuxemon or by their active tuxemon

So far we have only gone with the item used, the HP of the tuxemon and conditions affecting the tuxemon - and at the moment there is only one item, so that is effectively static.

Proposal

ShadowApex set up the following system:

The catch likelihood is compared to a random number between 0 and 1,000. If the result is lower than the catch likelihood, the capture is successful.

Catch Likelihood:

((HP Lost/Total HP) * 1,000) + 150 if affected by a condition + the Power of the item used (100 for a regular Capture Device)

The only change I would make is limiting the +150 to negative conditions.

Other options

Tension

In a lot of catching games, there's a pause before the result is declared - perhaps with the Capture Device shaking, beeping, etc., a number of times based on the probability of the capture succeeding.

This adds tension and excitement, but also provides feedback: if you keep getting just one beep before the Capture Device fails, your capture chances are low.

A simple way to do this is to take the square root, cube root or so on of the capture chance, and test that two, three or four times. If it succeeds each time, then the capture is successful.

For example, if a tuxemon has an 80% chance of being captured, and three successes are required, then the chance for each should be 93%. If the random number generator produces 13, 82 and 55, for example, the capture is successful, but if it produces 13, 82 and 99 it is unsuccessful - and if it produces 98 on the first result, it doesn't continue testing for success.

Capturing alternatives

There are a few other ways that capturing could work instead:

* One shot: Neo Monsters doesn't have any items in combat. Instead, you can attempt to capture each individual monster only once. This is a great way of creating tension: do I keep lowering their health further but risk knocking them out, or do I go now but have a greater risk of failure?
* Test of skill: In this case, you have to win a test of skill like playing a round of blackjack, or clicking a button at the right time. The more damage you do to the tuxemon/better item you use, the easier the test of skill is.
* Separate process: There could be a separate way of capturing tuxemon that is unrelated to battle. For example, defeated tuxemon could drop eggs in battle that can be hatched (like in Monster Sanctuary), or there could be a separate location you go to just for catching - no battling required.